Olfactory Cues in Mosquito Host Location

   
     Host-seeking mosquitoes are exposed to a wide variety of visual, olfactory, gustatory and physical 
stimuli.  Any one or combination of these stimuli could potentially act as cues for host identification and 
location.(3)   Visual and physical stimuli such as variations in skin temperature, moisture and skin color 
together with host odor stimuli provide the necessary cues for mosquitoes to locate their hosts.(5)  

     There has been extensive work done to determine the mechanism of mosquito attraction to its host.  
There is ample evidence that host seeking in mosquitoes is mediated by semiochemicals emanating from 
the host.  Olfactory cues are detected through an intricate pathway, beginning with sensilla located on the 
antennae which detect odor, and palpi which detect carbon dioxide.  Age and the physiological state of the 
mosquito determine whether the detection of olfactory cues results in a behavioral response.(24)  

      The behavioral role of odors released by mosquito hosts is poorly understood, indeed for many 
species it is still uncertain whether olfactory cues play a significant part in host location.(3)  Carbon 
dioxide induces a directed response serving to guide the mosquito towards the host.  The effect of carbon 
dioxide is a combined effect which increases the response to convection currents close to he host and the 
response to odor factors at a distance.(11) The purpose of this review article is to examine the mechanism 
for host selection by mosquitoes.  The synergistic role of carbon dioxide with L-lactic acid, and possibly 
with the bacterial fauna of the skin will be explored.  Physical factors relating to host selection will also be 
discussed.   The fact that people differ in attractiveness to mosquitoes will be discussed as well.  Is this 
difference due to the health of the host, or does the more attractive host have a different bacterial fauna on 
their skin?  Are there physical factors involved as well?
 
Role of Carbon Dioxide 

     Carbon dioxide is commonly used to attract mosquitoes to traps used for mosquito surveillance in field 
experiments.  The role of carbon dioxide in the host-seeking behavior of mosquitoes is a matter of debate.  
Some regard it as only an activator of flight behavior.(4)  I will be looking at the conclusions of Gillies 
who concludes that the role of carbon dioxide in host-seeking by mosquitoes is comprised of two distinct 
actions.  First, it acts as an attractant in which orientation towards the host is mediated by kinesis and 
optomotor anemotaxis.  Secondly, carbon dioxide can exhibit a synergistic response with host odors.(11)

Orientation/ attraction

     Carbon dioxide plays an essential role in the take-off behavior of the mosquito.  It has been 
demonstrated that in the absence of host stimuli, the take-off rate is essentially a random process.  When 
mosquitoes are exposed to an airstream to which .2% carbon dioxide is added, the rate of take-off is 
greatly increased for up to two minutes before falling back to a low level.  Similarly, it has been shown 
that carbon dioxide introduced at the base of a tower stimulated flying but not probing.(11)

     The orientating effect of carbon dioxide appears to be comparable to that of odor factors with one 
important difference.  The stimulus is effective when presented at a constant level.  However, only when 
the stimulus is pulsed do directed responses occur.(11)  Work done by Omer & Gillies demonstrated that 
that when carbon dioxide is presented as a pulsed stimulus into the airstream  at a rate of twenty seconds 
on and twenty seconds off, the mosquitoes move rapidly up the tunnel towards the inlet.  The conclusion 
can be made that carbon dioxide stimulates sustained flight if the following parameters are met.  First the 
level of the stimulus must be pulsed not constant,  and secondly the stimulus must be in a moving 
airstream.  These two stimuli combined result in upwind displacement by the mosquito.  These laboratory 
findings are in agreement with the results of releasing carbon dioxide in the field where the stimulus is 
received intermittently due to the turbulence of the wind.(11) 

     Carbon dioxide also acts as an attractant.  The response is initially one of activation followed by 
upwind flight.  The mosquito is being steered by optomotor responses elicited by ground pattern 
movement.  Thus orientation to the host is the result of two behavioral responses, kinesis and optomotor 
anemotaxis.  It is in this sense that carbon dioxide is acting as an attractant.(11)
 
Synergistic Properties of Carbon Dioxide     

     One of the most striking effects of carbon dioxide is to modify or augment the effects of other stimuli.  
Gillies(11) has concluded that in the case of carbon dioxide it appears that the combined action ranges from 
priming the response to a stimulus that by itself has no effect, through additive effects in combination with 
another stimulus, to a true synergism.  Experiments by Kahn et al. demonstrated that Aedes aegypti  will 
orient to the convection currents rising from a moist warm body.  This orientation is enhanced when 
carbon dioxide is added to the environment.(11,15)  The carbon dioxide is acting to enhance the attractive 
nature of host odors, or even combining with otherwise non-attractive volatiles resulting in increased 
attraction.

Lactic acid 

      Lactic acid is a volatile by product of anaerobic metabolism common to all animals and is present in 
human skin emanations.  Smith et el(22) in work with Aedes aegypti , found that carbon dioxide has a 
synergistic effect with lactic acid.  Charcoal filtered outdoor air was run through the olfactometer with 
10ug of lactic acid.  When the lactic acid was run by itself, no attraction was observed.  When the lactic 
acid was introduced in combination with carbon dioxide the mosquitoes were attracted.(22)  Further 
research by Eiras and Jepson(10) with Aedes aegypti has demonstrated that lactic acid has no effect at 
close range.  Olfactometer assays confirm that lactic acid has no effect on A. aegypti at close range, and 
seems to be most important in long distance host seeking.

       Lactic acid is detected by two classes of neurons.  These neurons are found within the grooved peg 
sensilla of mosquitoes that respond to lactic acid.(20)  One class of neurons is excited by lactic acid, while 
the other is inhibited at intensity ranges emanating from a human hand.(6)

     The presence or absence of host seeking behavior is correlated with the level of sensitivity of the lactic 
acid excited neuron.  Blood feeding in mosquitoes has been shown to be initiated between 24 and 72 hours 
after a female emerges.  A similar period of maturation appears to be required before the lactic acid 
receptors are fully responsive.(7)  In unpublished observations, Davis could not consistently obtain 
measures of receptor specificity and sensitivity unless the females were at least 3 to 4 days post 
emergence. 

Diapause  

     Adult diapause in mosquitoes is characterized by delayed reproduction, fat body hypertrophy, low 
metabolic rate and other behavioral and metabolic changes that enhance survival during inclement 
conditions brought about by seasonal change.  Culex pipiens females enter adult diapause in response to 
short day lengths experienced during larval and pupal development.  The follicles of diapausing adult 
females fail to undergo previtellogenic development and remain teneral until diapause is terminated.(1) 

      Female Culex pipiens  that have terminated adult diapause display changes in peripheral sensory 
response characteristics that differentiate them from diapausing females.  The changes reveal that diapause 
termination involves a renewal of peripheral receptor responsiveness that is related to the resumption of 
reproductive activity.  These changes occur in two distinct cell populations.  One type of receptor cell that 
undergoes changes are the sensilla type A2 which are sensitive to opviposition site attractants.  I will not 
discuss type A2 receptors further.  Sensilla basiconica type A3 cells are highly sensitive lactic acid excited 
cells that are present in females that have terminated diapause.(2)  The level of sensitivity of these type A3 
receptors is known to be correlated with the presence or absence of host-seeking behavior.

      The age of the mosquito has also been shown to be a factor in the sensitivity of the A3 receptors.  
Davis(7) has shown that Aedes aegypti do not exhibit host-seeking behavior before 18-24 hours post 
emergence.  However at 30 hours about 10% of the females tested began to exhibit host-seeking behavior.  
The 50% response level was reached at about 66 hours post emergence;  by 102 hours post emergence 
90% of the females were actively seeking a host.  The results of this study are illustrated in Figure 1.

     The females between 30 and 102 hours post emergence are in a transitional condition during which 
their host-seeking behavior is clearly age dependent.  The host-seeking behavior of virgin females of ages 
greater than 108 hours post emergence show a consistent response rate of 94% for as long as 15 days post 
emergence.(7)

Skin emanations 

     Volatile substances produced by human skin have been shown to act as either attractants or repellents.  
Many of the volatiles responsible for these actions are found in sweat.  There are three different types of 
human sweat glands, each of which produce slightly different blends of compounds.  The eccrine 
sudoriferous glands are distributed over the entire body, but are most abundant on the palms of the hands, 
forehead and the soles of the feet.  The apocrine sudoriferous glands are most numerous in the armpits, 
inguinal areas and around body aperatures.  Sebaceous glands are found most abundant on the face and 
scalp with none found on the palms of the hand or soles of the feet.(25)   Figure 2(25) illustrates the 
results of studies on mosquito olfactory responses to human sweat and skin emanations.  Samples of 
human sweat have been bioassayed by many workers with varying results.   Results range from early 
reports that sweat was unattractive to mosquitoes to observations that Aedes aegypti aggregate near sweat 
samples.  The major chemical component of attractive sweat was identified as lactic acid.  Lactic acid was 
confirmed to be an attractant for Aedes aegypti, but the general consensus was that there may be other 
chemicals in human sweat that are involved in host location.(22) The human hand elicits higher responses 
as compared to lactic acid alone or with either water vapor or temperature treatments.(11)  These results 
suggest that the human hand releases chemical stimuli other than lactic acid that are responsible for eliciting 
responses at close range.(10)   

     Between 300 and 400 compounds are constantly released as by products of metabolism.  Of these 
compounds released, approximately 200 of them are carboxylic acids.(23)  Many of these compounds are 
further modified into the equivalent alcohols, alpha-hydroxyacids and diols.  Odorous steroids have also 
been identified from apocrine glands associated with axillary, anogenital, sternal and areolar body 
regions.(3)   Although many of these emanations appear to be attractive, most attention has been given to 
lactic acid. The presence of well characterized host attractant receptors has lead to heightened interest.(1)  
As a result most research on host odors has been in relation to lactic acid. 

Temperature

     Temperature is an important physical signal associated with blood feeding arthropods.  Change in 
temperature can be detected and used to orient the organism to the host.  The thermosensitive neurons in 
mosquitoes are associated with the coeloconica sensilla at the tip of the antennae, and contain two types of 
thermosensitive neurons.  One of these receptors is a cold receptor and the other is a warm receptor, 
resulting in the neurons functioning to form a differential responding pair. Temperature changes on the 
order of 0.05 degrees Celsius  within convection currents arising from a 2-kg rabbit can be detected from 
more that 2 meters away.  Compression waves produced by a body moving in air 3 meters from the 
mosquito causes the thermoreceptors to respond with up to 30 impulses per second.(6)    

     Eiras and Jepson studied the response of Aedes aegypti to host odors in relation to convection currents.  
Female Aedes aegypti respond significantly to convection currents produced by the human hand.  The 
addition of water vapor to the convection currents enhanced significantly the response to the hand, while 
lactic acid alone or in combination with water vapor and convection currents did not increase the response 
level.  An extract of sweat elicited a higher response level than the convection currents as well as water 
vapor and lactic acid.  Human hands elicited the highest response which seems to suggest that there are 
other cues besides carbon dioxide, lactic acid and convection currents involved in host location.(10)

Selection of biting sites

     Blood feeding arthropods are generally not evenly distributed over their host's bodies.  Most approach 
their host by air, and usually feeding seems to occur on or close to the place where the insect lands.  When 
the host is an animal, foraging for an area free of thick hair usually occurs.  The effect of host hair on 
mosquito behavior is presumably limited for human hosts because less than 5% of the human skin surface 
is densely covered with hair.  Generally, biting on humans takes place on the landing site.(8) 

     The landing sites on humans have been demonstrated to be governed by several cues.  Heat and 
moisture convection currents have been reported to effect the biting behavior of mosquitoes.  Some species 
have been shown to only bite within a certain height above the ground.  Eretmapodites chrysogaster bites 
only between the ankles and knees of standing humans.  This behavior was shown not to be influenced by 
body heat or moisture since biting occurred all over the body when the host was lying down.(9,12)  Aedes 
simpsoni mainly bites on the heads of naked individuals.  Tests were done with the subject in standing, 
sitting and lying positions.  The results revealed that there was no influence of height above the ground on 
biting behavior.  The mosquito relies on body heat and moisture and probably other cues from the 
skin.(9,13)  These differences in biting site selection may reflect different host-seeking strategies.  
Eretmapodites chrysogaster, for example, does not actively search for a bloodmeal but waits until a host 
enters its visual range, a strategy based on visual host finding which implies a broad host spectrum.  
Opportunistic feeders may orient towards commonly produced chemical cues, whereas specialists would 
require host specific information.(9,12)

     De Jong and Knols(9) studied Anopheles atroparvus and Anopheles gambiae.  Each of these two 
species prefer different parts of the body.  A. atroparvus prefers the head region, while A. gambiae prefers 
the lower leg and foot region.  These preferences for these regions correlates with particular combinations 
of skin temperature and eccrine sweat gland density.  Interestingly, modification of the host odor profile 
by removing exhaled breath and washing the lower leg and foot region results in changes in these 
preferences.  

     Figures 3 and 4 illustrate the experiment.  Figure 3 shows the distribution of biting sites of the two 
species before the host profile was modified.  Figure 4 shows the distribution of biting sites after the host 
profile was modified.  Exhaled breath was removed from the room by using a one way breathing valve 
connected to polythene lay-flat tubing for the test with A atroparvus.  The host odor profile was changed 
for A gambiae by washing the ankles and feet with a non-perfumed medical soap containing a bacterial 
agent.(9)  

     This work by De Jong and Knols shows that different mosquito species exhibit different biting site 
preferences on one and the same human host and that preferences can be altered by changing the hosts 
body odor profile.  This clearly demonstrates that this process is not entirely governed by skin 
temperature, skin humidity and visual cues, but by body odors as well.(9)

Preferential biting

     Differential attractiveness of humans to mosquitoes has been documented.(18,19,21)  Although the 
mechanisms described earlier play a role in this phenomenon, most researchers agree that there are many 
more cues emanating from the human host that have yet to be described.  Recent experiments using a wind 
tunnel bioassay have suggested that human skin microflora might be responsible for producing 
compounds that attract mosquitoes.  Humans have been shown to have varying concentrations and species 
of skin bacteria resulting in the production of varying amounts of volatile substances that act as 
attractants.(14,17,16)   The ability to change preferences of mosquitoes by changing the host odor profile 
by washing the skin lends support to the role of bacteria in host-seeking.  No definite conclusions can be 
made on this interesting portion of the story until more research has been done.

Conclusion

    Although it is not clearly understood why some individuals are more attractive to mosquitoes than 
others, research shows that preferential biting does occur.  Preferential biting of certain individuals over a 
period of time is epidemiologically important because it demonstrates that some people within a community 
will be at a greater risk from mosquito borne pathogens than others.(19)  

     One might pose the question, why aren't host odors utilized in vector control strategies?  Currently 
available odor baited trapping systems rely on non-specific attractants such as carbon dioxide.  Despite the 
evidence suggesting that host odor cues, apart from carbon dioxide and moisture, are involved in the host 
seeking behavior of many anthrophilic mosquitoes, no such trapping systems utilizing these cues have 
been developed.  The complexity of human odor and the problems associated with distinguishing between 
the effects of human odor, temperature, humidity and carbon dioxide on mosquito behavior in the 
laboratory as well as in the field continue to hinder progress.(3)

     Recently, Knols and De jong have shown that limburger cheese acts as an attractant for the malaria 
mosquito Anopheles gambiae.  Many times carbon dioxide is not convenient for use in the field either from 
a gas cylinder or in the form of dry ice.  Limburger cheese acts as a attractant independently of carbon 
dioxide, and its discovery as an attractant for an important malaria vector is an important step foreword in 
the development of effective monitoring devices and possibly a means of control.  As work continues, one 
might foresee the development of odor baited traps which could be used both to enhance trapping 
effectiveness for epidemiological studies as well as provide a new means for control of mosquitoes.

     References

1.   Bowen, M.F.  1996.  Sensory aspects of host location in mosquitoes.  Ciba Foundation Symposia.  
200:197-211.

2.   Bowen, M.F.  1990.  Post-diapause sensory responsiveness in Culex Pipiens.  J. Insect Physiol.  
36(12):923-929. 

3.   Cork, A.  1996.  Olfactory basis of host location by mosquitoes and other haematophagous diptera.  
Ciba Foundation Symposia.  200:71-88.

4.   Costantini, C.,  Gibson, G.,  Sagnon, N.,  Della Torre, A.,  Brady, J.,  Coluzzi, M.  1996.  
Mosquito responses to carbon dioxide in a West African Sudan savanna village.  Med. Vet. Ent.  10:220-
227.

5.   Costantini, C.  1996.  Introduction III:  odors for host-finding mosquitoes.  Ciba Foundation 
Symposia.  200:124-131.

6.   Davis, E.E.,  Bowen, M.F.  1994.  Sensory physiological basis for attraction in mosquitoes.  J. Am. 
Mosq. Cont.  10(2):316-325.

7.   Davis, E.E.  1984.  Development of lactic acid receptor sensitivity and host-seeking behavior in newly 
emerged female Aedes aegypti mosquitoes.  J. Insect Pysiol.  30(3):211-215.

8.   de Jong, R.,  Knols, B.G.J.  1996.  Selection of biting sites by mosquitoes.  Ciba Foundation 
Symposia.  200:89-103.

9.   de jong, R.,  Knols, B.G.C.  1995.  Selection of biting sites by two malaria mosquito species.  
Experientia.  51:80-84.

10.  Eiras, A.E., Jepson, P.C.  1994.  Responses of female Aedes aegypti (Diptera: Culicidae) to host 
odors and convection currents using an ofactometer bioassay.  Bull. Ent. Res.  84:207-211.

11.  Gillies, M. T.,  1980.  The role of carbon dioxide in host-finding by mosquitoes (Diptera: Culicidae): 
a review.  Bull. Ent. Res.  70:525-532.

12.  Haddow, A.J.  1946.  36:33.

13   Haddow, A.J.  1956.  46:761.

14.  Keystone, J.S.  1996.  Of bites and body odor.  The Lancet.  347:1423.

15.  Khan, A.A.,  1966.  Quantitation of effect of several stimuli on landing and probing by Aedes 
aegypti.  J. Econ. Ent.  59:902-905.

16.  Knols, B.G.J., De Jong, R.  1996.  Limburger cheese as an attractant for the malaria mosquito 
Anopheles gambiae.  Parasit. Today.  12(4):159-161.

17.  Knols, B.G.J.,  De Jong, R., Takken, W.  1995.  Trapping system for testing olfactory responses of 
the malarial mosquito Anopheles gambiae in a wind tunnel.  Med. Vet. Ent.  8:386-388.

18.  Knols, B.G.J.  De Jong, R.,  Takken, W.  1995.  Differential attractiveness of isolated humans to 
mosquitoes in Tanzania.  Trans. R. Soc. Trop. Med. Hyg.  89:604-606.

19.  Lindsay, S.W.,  Adiamah, J.H.,  Miller, J.E.,  Pleass, R.J.,  Armstrong, J.R.M.  1993.  Variation 
in attractiveness of human subjects it malaria mosquitoes (Diptera: Culicedae) in the Gambia.  J Med. Ent.  
30(2):368-373.

20.  McIver, S.B.  1970.  Comparative study of antennal sense organs of female culicine mosquitoes.  
Can. Ento.  102:1258-1267.

21.  Schreck, C.E.,  Kline, D.L.,  Carlson, D.A.  1990.  Mosquito attraction to substances from the skin 
of different humans.  J. Am. Mosq. Cont. As. 6(3):406-410.

22.  Smith, C.N., Smith, N., Gouck, H.K., Weidhaas, D.E., Gilbert, L.H., Mayer, M.S., Smittle, B.J., 
Hofbauer, A.  1970.  L-lactic acid as a factor in the attraction of Aedes aegypti (Diptera: Culicidae) to 
human hosts.  AESA 63:760-770.

23.  Stoddart D.M.  1990.  The scented ape.  Cambridge University Press, Cambridge.

24.  Takken, W.  1996.  Synthesis and future challenges:  the response of mosquitoes to host odours.  
Ciba Foundation Symposia.  200:302-320.

25   Takken, W.  1991.  The role of olfaction in host-seeking of mosquitoes:  a review.  Insect Sci. 
Applic.  12:287-295.